Cytometric bead array immunoassay - M ATERIALS AND M ETHODS

3. M ATERIALS AND M ETHODS

3.8. Cytometric bead array immunoassay

The cytometric bead array (CBA) technique is a method that captures a set of chemical substances, known as analytes, with beads of known size and fluorescence, making it possible to detect analytes using flow cytometry.

For the CBA measurements, the BD™ CBA Mouse Inflammation Kit (BD Biosciences, California, USA) was used to quantitatively measure Interleukin-6 (IL-6), Interleukin- 10 (IL-10), Monocyte Chemoattractant Protein-1 (MCP-1), Interferon-γ (IFN-γ), Tumor Necrosis Factor (TNF), and Interleukin-12p70 (IL-12p70) protein levels in a single sample. The assay was performed according to an adjusted manufacturer’s protocol.

Briefly, each capture bead conjugated with a specific antibody was mixed in a single tube.

The standard curve for each protein covered a defined set of concentrations from 0 to 5000 pg/mL. Inside the 96-well plate, 10μl of capture beads mixture was added to the same amount of recombinant standards or unknown samples and then incubated in the dark with 10μl of phycoerythrin (PE) - conjugated antibodies at RT for 2h. This PE detection reagent provides a fluorescent signal in proportion to the concentration of a specific cytokine, which is quantified from a calibration curve. Posteriorly, each well was washed with wash buffer and the plate was centrifuged at 400g for 5 minutes. After discarding the supernatant, each pellet was resuspended.

These sandwich complexes formed by capture bead plus analyte plus detection reagent were measured using a BD™ FACS Calibur flow cytometer (BD Biosciences) either through the 96-well plate or 12 × 75-mm tubes. Each tissue lysate or medium sample was quantified three times in different days.

Data were obtained and analyzed by FCAP Array software in dot-plot FL-2 channel vs. FL-3 channel. FL-2 detects PE, which emits at 585 nm, and FL-3 detects the particles that were dyed with six different fluorescence intensities and has a maximal emission wavelength of approximately 650 nm. The six FL-3 particles dyed to different intensities were distributed along the y-axis. The concentration of the cytokine calibrators was expressed (y-axis) vs.

medium fluorescence intensity (FL-2) in the standard curves. The concentrations of cytokines that were below the limit of detection of the assay were given zero value of concentration.

25 3.9. Immunohistochemistry

Immunohistochemistry demonstrates the presence and location of proteins in an intact tissue. In this sensitive technique, a primary antibody recognizes specifically the target protein.

This interaction can be later detected by a fluorochrome-conjugated secondary antibody and visualized using fluorescence microscopy.

Slices were fixed after treatment at 15 DIV. The NBA culture medium was removed and 1mL of 4% paraformaldehyde (PFA; VWR, Pennsylvania, USA) diluted in PBS (137mM NaCl, 2.1mM KCl, 1.8mM KH2PO4 and 10mM Na2HPO4.2H2O, at pH 7.40) was added beneath and above the slices in each well of the culture tray. Later, slices were sequentially incubated for 1h in increasing concentrations of sucrose in PBS (10%, 20% and 30%), to cryoprotect the tissue.

Slices were maintained (for a maximum of one week) at 4ºC soaked in 30% sucrose until further use.

In the beginning of the immunohistochemical staining procedure, the insert membrane was cut with a scalpel to individualize each slice and then each slice was transferred to a microscope slides (Thermo Fisher). In each slide, two individual slices were surrounded by a hydrophobic pen, DAKO pen (Dako, Glostrup, Denmark), to protect slices from drying out. Also to prevent the tissue from drying out, all incubations were carried out in a humidified chamber. Drying at any stage promotes non-specific binding and high background staining.

After 3 times of PBS washes, 10min each, slices were permeabilized with 1% Triton X-100 (Aldrich) in PBS for 1h at RT and then blocked with 20% Donkey Serum (D9663, Sigma-Aldrich) and 1% BSA in Milli-Q water for 3h at RT. Serum raised in donkey was chosen since it was the host of all secondary antibodies, thus minimizing the cross reactions with endogenous immunoglobulins in the tissue. BSA was also included to reduce non-specific binding caused by hydrophobic reactions. Primary antibodies (Table 3) diluted in PBS were applied and incubated overnight at 4ºC.

In the ensuing day, after washed 3 times for 10min with PBST (PBS with 0.1% Tween-20), slices were incubated with the fluorophores-labeled secondary antibodies (Table 4) in PBS for 5h at RT. Following rinse again with PBST (3 times) in the dark to avoid photobleaching, DAPI (D9564, Sigma-Aldrich) solution (1:1000 dilution in PBS) was added over 40 min at RT to stain the nuclei. To finalize, slices were rinsed 3 times with PBST and one with PBS and then mounted in Mowiol Solution (2,4g Polyvinylalcohol 4-88, 600 mM glycerol, 200mM tris, pH 8.0, in Milli-Q water). It is noteworthy to state that in the wash, the sections were wiped around with a tissue paper to drain the excess liquid.

Slices were observed in a confocal laser scanning microscope (Zeiss LSM 710, Carl Zeiss NLRP3 staining), a threshold below the most saturated pixels was applied. The percentage of co-localization was calculated by the area of NLRP3 saturated pixels which co-localized with GFAP divided by the area of the all saturated pixels in a confocal image. The threshold was set at 100 for the red channel and at 30 for the green channel.

3.10. Electrophysiology – extracellular recordings

At 15-17 DIV, slices were removed from the incubator and placed in a petri dish with heated NBA medium. To individualize one slice at a time, the insert membrane was cut with a scalpel and the slice was transferred to an interface recording chamber with a humidified 95%

O2/5% CO2 atmosphere at 37° C. Unlike the immersion chamber, this interface chamber allows the medium to pass under the slice as it was in the incubator. The NBA medium was superfused and recirculated at a rate of 2mL/min.

Recordings were obtained with an Axoclamp 2B amplifier (Axon Instruments, Foster City, California, USA), digitized and continuously stored on a personal computer with the WinLTP software (WinLTP Ltd., Bristol, UK)(Anderson and Collingridge, 2001). The pCLAMP Software Version 10.7 (Molecular Devices Corporation, California, USA) was used for data analysis. All recordings were band-pass filtered (eight-pole Bessel filter at 60 Hz and Gaussian filter at 600 Hz).

As the entorhinal cortex is included in these slices and its projections are intact, ictal discharges are originated in entorhinal cortex and propagate through DG to CA3 and CA1. Also, interictal discharges initiate in the CA3 and propagate via CA1 and subiculum to the entorhinal cortex and return to the hippocampus through the DG (Barbarosie and Avoli, 1997; Rutecki and Yang, 1998; Walther et al., 1986). As discharges can lose power in CA1 and this hippocampal region has more percentage of neuronal death in OHSC, the extracellular recordings to monitor the electrical activity of neurons were made in CA3 pyramidal cell region (Fig. 8). The viability of slices was routinely tested by recording population spikes from CA3 pyramidal cell population (Fig. 8).

3.10.1. Population spikes

Population spike is the synchronous discharge of neuron populations (Dyhrfjeld-Johnsen et al., 2010). Mossy fiber projections to CA3 pyramidal cells were electrically stimulated for the purpose of recording a biological response in the form of a population spike.

Extracellular recordings of population responses in the CA3 area were made using glass micropipettes electrode (2–4 MΩ) filled with artificial cerebrospinal fluid (ACSF) composed by:

124mM NaCl, 3mM KCl, 1.2mM NaH2PO4, 25mM NaHCO3, 10mM glucose, 2mM CaCl2, 1mM MgSO4 with pH 7.4.

A bipolar concentric wire stimulating electrode was placed on mossy fibers and rectangular pulses of 0.1ms duration at every 15s were evoked. The average of 8 consecutive population spikes was obtained for representative purposes. Slices were stimulated with intensity between 1 and 4 volts. If there was no response, the recording or stimulating electrode was repositioned. After this second attempt if there was still no response the slice was discarded.

WinLTP 2.20b Reanalysis software (WinLTP Ltd., Bristol, UK)(Anderson and Collingridge, 2001) was used to visualize the population response.

Fig. 8 - Extracellular recording of one

27 3.10.2. Epileptiform activity

Extracellular recordings of epileptiform activity in the CA3 pyramidal cell region were performed using the same glass micropipette electrode used to record the population response, without changing the place (Fig. 8). The activity was recorded for at least 40 minutes.

In this study, interictal epileptiform discharges were defined as individual paroxysmal discharges clearly distinguished from the baseline, with an abrupt change in polarity occurring within several milliseconds (Berdichevsky et al., 2012). Ictal-like discharges were defined as continuous discharges lasting more than 6s (bursts). The end of a burst was defined when inter-spike interval was longer than 600ms. Continuous spike activity with duration lower then 6s was not accounted as burst activity. Slices with this type of activity and interictal discharges were classified as slices with interictal activity.

To quantify the epileptiform activity different parameters were assessed. The number of burst per slice was evaluated manually in slices with interictal-like discharges and in slices with mixed interictal-like and ictal-like discharges. To further characterize the bursts of the latest slices, the frequency of events within burst and the average positive peak amplitude (this is the amplitude between the baseline and the peak of the spike) was evaluated. The number of events per burst, the duration of each burst and the positive peak amplitude were auto-detected by pCLAMP Software, which allows a more reliable and automatic detection of events. The baseline used to detect these events was specific to each recording and was settled right above the end of noise oscillations.

3.10.3. Experimental conditions recorded in electrophysiology

In a first set of experiments, recording conditions included the drug free control condition (CTL), the inflammasome activation condition (LPS/ATP) and the inflammasome inhibition condition in an inflammatory context (MCC950/LPS/ATP). As negative controls, slices incubated with LPS (10ng/mL) alone over 3h or with ATP (1mM) alone over 3h were also evaluated (Fig. 9).

After these initial electrophysiological recordings, it was hypothesized that the change in culture medium on the day before the treatment was an inflammatory trigger of epileptiform activity. Therefore, studies were performed on slices maintained in Opti-MEM medium, without undergoing the change to NBA, and on slices exposed only to MCC950, incubated in NBA for 7h (Fig. 9). It should be noted that the neuronal electrical activity of the slices was recorded on time or 1h after the end of incubation, since more than one slice per insert was recorded.

28 3.11. Drugs and antibodies

Table 2 - Drugs used for slices treatment.

Drugs Vehicle Stock

concentration

Final

concentration Reference Supplier Lipopolysaccharide (LPS) Milli-Q

water 5mg/ml 5, 10 or 20

ng/mL L6529

Sigma-Aldrich Adenosine

5′-triphosphate disodium salt hydrate (ATP)

NBA 50mg/mL 1 mM A7699

Sigma-Aldrich

MCC950 DMSO 10mM 10 μM 17510

Cayman chemical company Dimethyl sulfoxide

(DMSO) - Density:

1.10 g/mL 0.1% D2650

Sigma-Aldrich

Table 3 - Primary antibodies used in western blot and immunohistochemistry assays.

Protein Antibody Technique and

Dilution Reference Supplier

αII-Spectrin Mouse monoclonal WB - 1:500 sc-48382 Santa Cruz Biotechnology ASC Rabbit polyclonal WB - 1:1000 AL177 AdipoGen Life Sciences GAPDH Mouse monoclonal WB - 1:1000 AM4300 Invitrogen

GFAP Rabbit polyclonal WB - 1:5000 G9269 Sigma-Aldrich GFAP Mouse monoclonal IHC - 1:500 MAB360 EDM Milipore

Iba1 Goat polyclonal WB - 1:1000 ab5076 Abcam

Iba1 Rabbit polyclonal IHC – 1:250 ab108539 Abcam

NLRP3 Rabbit polyclonal WB - 1:300

IHC - 1:500 ab214185 Abcam

Fig. 9 - Schematic representation of the conditions used in electrophysiology.

Table 4 - Secondary antibodies used in western blot and immunohistochemistry assays.

Secondary Antibodies Technique

and Dilution Reference Supplier

Donkey anti-Goat IgG -HRP WB - 1:10000 sc-2020 Santa Cruz Biotechnology Goat anti-Mouse IgG -HRP WB - 1:10000 sc-2005 Santa Cruz Biotechnology Goat anti -Rabbit IgG -HRP WB - 1:10000 1706515 Bio-Rad

Donkey anti-Rabbit Alexa Fluor® 488 IHC - 1:500 A21206 Invitrogen Donkey anti-Mouse Alexa Fluor®

488 IHC - 1:500 A21202 Invitrogen

Donkey anti -Rabbit Alexa Fluor®

568 IHC - 1:750 A10042 Invitrogen

3.12. Statistical Analysis

All statistical analyses were performed with GraphPad Prism 6.0. Statistical significance was determined by using one-way analysis of variance (ANOVA) followed by Tukey’s test for multiple comparisons, with p<0.05 considered to represent statistical significance. Data were expressed as means ± standard error of mean (SEM), except when n=2.

30 4. RESULTS

4.1. Establishment of NLRP3-mediated inflammation model

LPS is a potent bacterial endotoxin widely used to establish inflammation models. LPS is an excellent trigger of inflammation either in CNS or peripheral nervous system (PNS) and it has been used in different models, both in vitro and in vivo. Moreover, LPS was already described as an enhancer of seizure susceptibility (Sayyah et al., 2003) and an activator of NLRP3 inflammasome in in vivo models (F Zhang et al., 2016; Z-T Zhang et al., 2016).

In our study we aimed to establish an inflammatory phenotype in organotypic slices, but without damaging extensively the slices. Inflammation induction was tested with several LPS concentrations, ranging from 5 to 20ng/ml, over two-time periods, 3h and 6h. Furthermore, because NLRP3 inflammasome activation is described as a two-step process which requires two triggers, an extra condition which included ATP, was added to ensure that NLRP3 inflammasome was activated in our system.

Therefore, at 14 DIV, organotypic slices were exposed to 5, 10 and 20ng/mL of LPS over 3h and 6h, or to 10ng/mL of LPS alone for 3h followed by 3h co-incubation with 1mM of ATP.

4.1.1. Cell death assessment

Cell death was characterized through the expression of αII-Spectrin, a structural protein of the cell cytoskeleton mainly found in neurons (J He et al., 2016; Riederer et al., 1986). αII-Spectrin is a substrate for two cysteine proteases: calpains, which are related with necrosis and excitotoxicity, and caspase-3, which is the main effector caspase for apoptosis. Spectrin catabolism products are known as spectrin breakdown products (SBDP) and have distinct molecular sizes according to the enzyme responsible for the cleavage. Proteolysis by calpains forms fragments with 145kDa (SBDP145), whereas proteolysis by caspase-3 produces fragments with 120kDa. Another fragment with 150kDa (SBDP150) is produced by both proteases (Z Zhang et al., 2009).

Western blot analysis of tissue lysates from slices exposed to different LPS concentrations and to different time of exposure showed no differences in αII-Spectrin expression within conditions (Fig. 10A and B). Absence or presence of ATP also did not influence αII-Spectrin expression in these slices (Fig. 10C).

The effect of these different conditions in calpain-mediated necrosis was assessed through the ratio between SBDP145 and αII-Spectrin. It is important to note that SBDP150 and SBDP145 were analyzed as a single fragment. There were no differences in necrosis within slices exposed to the different concentrations of LPS and distinct exposure time (Fig. 10D).

However, ATP significantly decreased the ratio SBDP145/αII-Spectrin (0.8925±0.1586, n=4), when compared with slices incubated with the same concentration of LPS but in the absence of ATP (1.368±0.1692, n=4, ^#p<0.05, Fig. 10E). This decrease can be due to the occurence of another form of cell death, called pyroptosis, in slices exposed to LPS/ATP, which is closely related with NLRP3 inflammasome activation and release of IL-1β (Bergsbaken et al., 2010).

Regarding the effect of LPS in different conditions or in presence of ATP in caspase-3-mediated apoptosis, the ratio SBDP120/αII-Spectrin remained unchanged between conditions (Fig. 10F and G).

G αII-Spectrin (240KDa)

Fig. 10 - Expression profiles of αII-Spectrin and SBDPs in organotypic cortex-hippocampus slices following LPS exposure in presence or absence of ATP. (A) Representative immunoblots for αII-Spectrin, SBDPs and GAPDH after 3h or 6h of LPS incubation at different concentrations, and co-incubated with ATP (1mM). (B-G) Western blot analysis of αII-Spectrin [n=2-6] (B), ratio SBDP145/αII-Spectrin [n=2-6] (D) and ratio SBDP120/αII-Spectrin [n=2-4]

(F) of different concentrations of LPS in distinct timepoints, or co-incubated with ATP [n=4-5] (C), [n=4] (E), [n=2-4]

(G), respectively. GAPDH was used as the loading control. Data are presented as mean ± SEM (except when n=2).^#p<0.05 LPS vs LPS+ ATP by one-way ANOVA followed by Tukey’s test.

32 4.1.2. Expression of inflammatory markers

4.1.2.1. NLRP3/ ASC

NLRP3 inflammasome is a multiprotein complex formed by three distinct proteins: NLRP3 itself, ASC and pro-caspase-1. NLRP3 and ASC are the first to assembly together and subsequently recruit pro-caspase-1 (Hoss et al., 2017). Therefore, to evaluate NLRP3 inflammasome activation, the expression of NLRP3 and ASC subunits were assessed.

Contrariety to expected, there were no statistical differences in NLRP3 and ASC expression, neither between different LPS concentrations at 3h or 6h (Fig. 11B and Fig. 12B, respectively), nor between presence or absence of ATP (Fig. 11C and Fig. 12C, respectively). However, although the results do not confirm the NLRP3 inflammasome activation, they are also not sufficient to sustain that the inflammasome is not activated. Moreover, if we look carefully to the western blotting representative bands (Fig. 11A) it is possible to observe a similar expression of control conditions when compared with the others. Data described elsewhere has shown less NLRP3 expression in basal conditions (Meng et al., 2014; F Zhang et al., 2016).

B C

Fig. 11 - Expression profile of NLRP3 in organotypic cortex-hippocampus slices following LPS exposure in presence or absence of ATP. (A) Representative immunoblots for NLRP3 and GAPDH after 3h or 6h of LPS incubation at different concentrations, and co-incubated with ATP (1mM). Western blot analysis of NLRP3 expression in slices exposed to different concentrations of LPS at 3h and 6h [n=6-8] (B) and in presence of ATP [n=7-8] (C). GAPDH was used as the loading control. Data are presented as mean ± SEM.

4.1.2.2. Iba1/GFAP

Under dangerous or pathogenic conditions, glial cells are the key mediators of neuroinflammation. Therefore, activation of microglia and astrocytes is considered a hallmark of neuroinflammation. Changes in the activation state of microglia and astrocytes are known to increase the expression of their markers, Iba1 and GFAP, respectively (Bederson et al., 2001;

Ben Haim et al., 2015). Thus, a western blotting analysis was performed to evaluate Iba1 and GFAP expression within the various conditions tested.

After 3h or 6h of exposure to different concentrations of LPS, slices did not have alterations in Iba1 or GFAP expression in relation to control condition (Fig. 13B and Fig. 14B).

Furthermore, by comparing slices incubated with LPS in absence or presence of ATP, there were also no differences in expression of glial cells markers (Fig. 13C and Fig. 14C). Similarly to NLRP3 and ASC discussed in the previous section, Iba1 and GFAP expression in control condition was identical to the other conditions (Fig. 13A and Fig. 14A). In summary, the conditions tested did not induce an upregulation in microglia and astrocytes markers.

B C

Fig. 12 - Expression profile of ASC in organotypic cortex-hippocampus slices following LPS exposure in presence or absence of ATP. (A) Representative immunoblots for ASC and GAPDH after 3h or 6h of LPS incubation at different concentrations, and co-incubated with ATP (1mM). Western blot analysis of ASC expression in slices exposed to different concentrations of LPS at 3h and 6h [n=3-5] (B) and in presence of ATP [n=3-5] (C). GAPDH was used as the loading control. Data are presented as mean ± SEM.

C A

B A

Fig. 13 - Expression profile of Iba1 in organotypic cortex-hippocampus slices following LPS exposure in presence or absence of ATP. (A) Representative immunoblots for Iba1 and GAPDH after 3h or 6h of LPS incubation at different concentrations, and co-incubated with ATP (1mM). Western blot analysis of Iba1 expression in slices exposed to different concentrations of LPS at 3h and 6h [n=2-6] (B) and in presence of ATP [n=4-6] (C). GAPDH was used as the loading control. Data are presented as mean ± SEM (except when n=2).

Fig. 14 - Expression profile of GFAP in organotypic cortex-hippocampus slices following LPS exposure in presence or absence of ATP. (A) Representative immunoblots for GFAP and GAPDH after 3h or 6h of LPS incubation at different concentrations, and co-incubated with ATP (1mM). Western blot analysis of GFAP expression in slices exposed to different concentrations of LPS at 3h and 6h [n=3-7] (B) and in presence of ATP [n=4-7] (C). GAPDH was used as the loading control. Data are presented as mean ± SEM.

35 4.1.2.3. IL-1β production

Il-1β is the main pyrogenic product from NLRP3 inflammasome activation. It is well established that LPS through Nf-kB upregulates Il-1β gene resulting in the increase of pro-Il-1β, the inactive form. In the presence of ATP, the second stimulus, caspase-1 is activated to process pro-Il-1β, forming an active and mature form which can be released into the medium.

Therefore, to further assess the activation of NLRP3 inflammasome, Il-1β levels were quantified.

Fig. 15A shows that hippocampal Il-1β levels increased with increasing concentration of LPS.

Slices exposed for 3h to 10ng/mL (3229±721.0, n=3, *p<0.05) or 20ng/mL (3504±663.04, n=3,

*p<0.05) of LPS had higher expression of this cytokine when compared with control slices (371.8± 162.0, n=3). Similarly, slices exposed for 6h to any concentration of LPS (5ng/mL:

4964±354.0, n=3; *p<0.05; 10ng/mL: 6038±444.4, n=3; *p<0.05; 20ng/mL: 11824±1826, n=5;

****p<0.0001) presented significantly higher Il-1β levels when compared with control (305.9±63.0, n=7). Moreover, exposure to 20ng/mL over 6h induces the highest levels of Il-1β

No documento NLRP3 inflammasome as a target to reduce epileptiform activity in organotypic slices (páginas 44-0)